Post on 01-Mar-2020
EUR 23056 EN - 2007
VALIDATION OF AN ANALYTICAL METHODTO DETERMINE THE CONTENT OF
SUCRALOSE IN BEVERAGESREPORT ON A METHOD VALIDATION BY COLLABORATIVE STUDY
I. DONCHEVA and JOERG STROKA
The mission of the IRMM is to promote a common and reliable European measurement system in support of EU policies. European Commission Joint Research Centre Institute for Reference Materials and Measurements Contact information Address: Retieseweg 111, 2440 Geel, Belgium E-mail: Joerg.Stroka@ec.europa.eu Tel.: +32 14 571 229 Fax: +32 14 571 783 http://irmm.jrc.ec.europa.eu/ http://www.jrc.ec.europa.eu/ Legal Notice Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use which might be made of this publication.
Europe Direct is a service to help you find answers to your questions about the European Union
Freephone number (*):
00 800 6 7 8 9 10 11
(*) Certain mobile telephone operators do not allow access to 00 800 numbers or these calls may be billed.
A great deal of additional information on the European Union is available on the Internet. It can be accessed through the Europa server http://europa.eu/ JRC 42362 EUR 23056 EN ISSN 1018-5593 Luxembourg: Office for Official Publications of the European Communities © European Communities, 2007 Reproduction is authorised provided the source is acknowledged Printed in Belgium
VALIDATION OF AN ANALYTICAL METHOD TO DETERMINE THE CONTENT OF SUCRALOSE IN
BEVERAGES
REPORT ON A METHOD VALIDATION BY COLLABORATIVE STUDY
Determination of Sucralose in Beverages by
Thin Layer Chromatography in Combination with Reagent-free Derivatisation and
Ultraviolet/Fluorescence Detection
I. DONCHEVA and J. STROKA
European Commission, Joint Research Centre, Institute for Reference Materials and Measurements, Geel, Belgium
Contents
Section Page
Abstract............................................................................................................................3 Introduction .....................................................................................................................5 Test materials for the collaborative study........................................................................6 Split level samples ...........................................................................................................8 Organisation of collaborative study.................................................................................9 Results and Discussion ..................................................................................................11 References .....................................................................................................................16 Annex….………………………………………………………………………………17
- 3 -
Abstract
An inter-laboratory comparison was carried out to evaluate the performance
characteristics of a method for the determination of sucralose in beverages, which was
developed at the JRC-IRMM. The method is based on high-performance thin layer
chromatography (HPTLC), and reagent-free derivatisation followed by
ultraviolet/fluorescence detection. It was tested for the determination of Sucralose
(C12H19Cl3O8; (2R,3R,4R,5S,6R)-2-[(2R,3S,4S,5S)-2,5-bis(chloromethyl)-3,4-
dihydroxy-oxolan-2-yl]oxy-5-chloro-6-hydroxymethyl)oxane-3,4-diol; CAS No:
56038-13-2) in carbonated and still alcoholic and non-alcoholic beverages at proposed
European regulatory limits according to Directive 2003/115/EC (1). Precise
determination of Sucralose levels in some of the matrices for which European
legislative limits apply, required a robust and reliable analytical method. HPTLC
employing reagent-free derivatisation offered such a reliable but simple, fast, cost-
effective and environment friendly method (very limited quantities of organic solvents
methanol and acetonitrile were used). Separation of Sucralose was performed by direct
application of samples (diluted, degassed and/or filtered, if necessary) on amino-
bonded silica gel HPTLC plates without prior cleanup and development with
acetonitrile:water. The sweetener was determined after heating of the developed plate
to 190°C and quantified both in ultraviolet absorption and fluorescence measurement
mode. Beverages spiked with sucralose as well as beverages taken from the market and
labelled to contain sucralose, were sent to 14 laboratories in five different countries
following IUPAC guidelines. A sample that did not contain measureable amounts of
sucralose was spiked at levels of 30.5 mg/L, 100.7 mg/L and 299 mg/L. Recoveries
ranged from 104 – 125 % with an average of 112 % for ultraviolet detection and from
98 – 101 % with an average of 100 % for fluorescent detection. Based on results for
spiked samples (blind duplicates at three levels), as well as samples containing
Sucralose (blind duplicates at three levels and one split level), the relative standard
deviation for repeatability (RSDr) ranged from 10 – 31 % for ultraviolet detection and
from 9 – 16 % for fluorescence detection. The relative standard deviation for
reproducibility (RSDR) ranged from 14 – 31 % for ultraviolet detection and from 9 –
21 % for fluorescence detection. The limit of quantification on the basis of 10x the
baseline noise was 6 mg/L and response was linear in the range between 30 – 150
- 4 -
ng/spot. The method is therefore considered suitable for the determination of Sucralose
in beverages at the proposed European legislative limits.
- 5 -
Introduction
Sucralose was approved in the EU as a sweetener in food products according to
Directive 2003/115/EC (1). It is a non-volatile substance, which does not contain a
chromophore to facilitate detection. Methods elaborated for its determination involved
high performance liquid chromatography with refractive index detection (2), high
performance anion exchange chromatography with pulsed amperometric detection (3),
and capillary electrophoresis with indirect ultraviolet absorbance (4). Another
approach by high performance liquid chromatography with evaporative light scattering
detection for the determination of multiple sweeteners, including Sucralose, had been
described (5). That method was recently collaboratively tested. A promising alternative
was the simple and fast high performance thin layer chromatographic method (6). It
required very little or no sample preparation. Sucralose was separated on amino-
HPTLC plates and heated at 190° C. At that temperature the substance reacted with the
amino groups of the HPTLC layer (reagent free derivatisation) to form a brilliant
yellow coloured spot of unknown structure that could be measured both in ultraviolet
absorption and fluorescence mode. The method allowed Sucralose determination at the
levels of interest regarding European legislation.
- 6 -
Test materials for the collaborative study
For this inter-laboratory comparison exercise the following products were purchased
from food supermarkets or kindly provided by Sucralose producer Tate & Lyle
Specialty Sweeteners (Reading, United Kingdom): various brands of energy-reduced
water based flavoured drinks, juice drinks, spirit drinks containing less than 15%
alcohol by volume, and food supplements in liquid form. All were labelled to contain
Sucralose. Coke® Light was used as a blank material.
The type of the test materials is given in Table 1.
Table 1: Type of test materials
Test Material Type Blank Energy-Reduced Beverage
Low Level Energy-Reduced Food Supplement Medium Level 1 Energy-Reduced Flavoured Water Medium Level 2 Energy-Reduced Cloudy Fruit Juice
High Level Energy-Reduced Low Alcohol Spirit Drink
Original products were degassed, if necessary, and similar products were mixed to
obtain the levels as stated in Table 2.
Table 2: Target levels of Sucralose, mg/L, in test samples
Test Material Sample Sucralose, [mg/L] Low Level (Blind Duplicates) 04; 08 42 Medium Level 1 (Split Level) 01; 03 69
Medium Level 2 (Blind Duplicates) 02; 06 105 High Level (Blind Duplicated) 05; 07 112
As spiking by each participant would require dispatching large volumes of blank
material, the procedure was performed at the Institute of Reference Materials and
Measurements. The Spiking Protocol can be found in the Annex.
Spiked levels are shown in Table 3.
- 7 -
Table 3: Levels of Sucralose, mg/L, in spiked samples
Test Material Sample Sucralose, [mg/L] Low Level A; E 30.5
Medium Level B; D 100.7 High Level C; F 299
The materials were subsequently filled into 4 mL glass vials with screw caps (3.2 mL
in each vial) and kept at 4° C until dispatch for collaborative trial testing.
- 8 -
Split level samples
The two split level test materials (Sample 01 and Sample 03), nearly identical but of
slightly different composition (≤5% difference in composition), were obtained by
diluting one portion of the material (Sample Medium level 1) with a small amount of
diluent (5% of water by volume). Both portions were supplied to the participating
laboratories as test samples, each under a random code number, and each test sample
was analyzed only once. These 2 test samples constitute a split level sample and should
be considered Youden matched pairs (YP).
Table 4: Sucralose results from analysis of split level test material
Measurement
Mode
Sample Sucralose,
[mg/L]
yc ≥ 0.95xc
YES/NO
Sample 01 73 UV
Sample 03 71 YES
Sample 01 69 FL
Sample 03 66 YES
Results from Table 4 identify Sample 01 and Sample 03 as a Youden matched pair
(YP).
- 9 -
Organisation of the collaborative study
The instructions for participants in the inter-laboratory comparison are given in the
Annex of this report. A total of 14 collaborators from five different countries were
invited to participate in the collaborative trial. 13 participants returned valid results. All
names (in alphabetic order) and addresses of the participants are given in Table 5.
Table 5: List of participants who returned results (in alphabetic order) in the inter-laboratory comparison exercise for the determination of Sucralose
Participant Institution Address
A. Koch FROHME-APOTHEKE Frohmestrasse 14, D-22457 Hamburg, Germany
B. Caemmerer TU Berlin, Institut für Lebensmittelchemie Gustav-Mayer-Allee 25, D-13355 Berlin, Germany
B. Spangenberg University of Applied Science Offenburg, Fachbereich Verfahrenstechnik Badstrasse 24, D-77652 Offenburg, Germany
G. Morlock University of Hohenheim Garbenstrasse 28, D-70599 Stuttgart, Germany
HHAC HHAC Labor Dr. Heusler GmbH Hindenburgstrasse 33, D-76297 Stutensee, Germany
J. Große-Damhues Gemeinsames Chemisches und
Lebensmitteluntersuchungsamt für den Kreis Recklinghausen (CEL)
Kurt-Schumacher-Allee 1, D-45657 Recklinghausen, BGermany
J. Kemme Landesbetrieb Hessisches Landeslabor, Standort Kassel Druseltalstraße
Druseltalstrasse 67, D-34131 Kassel, Germany
K. Bouten JRC-IRMM Retieseweg 111, B-2440 Geel, Belgium
M. Schulz Merck KGaA Frankfurter Strasse 250, D-64293 Darmstadt, Germany
M. Vega University of Concepcion, Faculty of Pharmacy
Barrio Universitario, CL-4089100 Concepcion, Chile
R. Schneider CVUA Karlsruhe Weissenburgerstrasse 3, D-76187 Karlsruhe, Germany
T. Dzido Medical University of Lublin Staszica 6, PL-20-081 Lublin, Poland
V. Widmer, M. Steiner CAMAG Laboratory Sonnenmattstrasse 11, CH-4132 Muttenz, Switzerland
- 10 -
Prior to the trial, each participant received a questionnaire to evaluate the type of
equipment available. Then a pre-trial was conducted to familiarize the participants with
the method and to report back any problems experienced. The pre-trial involved the
analysis of a blank and a spiked sample. Remarks and comments were received from
seven participants, the most important ones being (a) to use 10 mL instead of 5 mL
volumetric flasks for diluting the test material, (b) to mention explicitly in the
procedure that the TLC chamber should not be saturated (provided a chamber for TLC
and not for HPTLC is used), (c) to put a clear sign on the parcel with the samples upon
receipt to be stored in a refrigerator. These comments were taken into account for the
trial. In addition, four participants observed cracks in the layer of the HPTLC plates
just after delivery. As a result additional shipping precautions were taken to ensure safe
shipment of the NH2-HPTLC glass plates.
For the collaborative trial each participant received:
1. Eight coded test materials (blind duplicates at three concentration levels and
one split level).
2. One vial marked ‘Sucralose Standard’ containing Sucralose, which was to be
employed as the calibrant, as described in the method.
3. Six vials marked ‘Spike solution A, B, C, D, E, and F’.
4. Eight NH2-HPTLC glass plates, 10 x 10 cm.
5. Five HPLC syringe filters, 0.45 µm.
6. One 3 mL syringe.
7. A copy of the collaborative study method.
8. A ‘Collaborative Study Materials Receipt’ form.
9. Report forms.
Each participant was required to dilute the test solutions and the spiked ones and to
perform a single analysis of each material by HPTLC.
Method of analysis
The method of analysis that was used in this study can be found in the Annex.
- 11 -
Results and Discussion
Collaborative trial results
All data submitted for the study is presented in Table 6 & Table 7. The data is given as
individual pairs of results for each laboratory (identified by the laboratory
identification codes). Blank samples spiked at levels of 30.5 mg/L, 100.7 mg/L, and
299 mg/L of Sucralose are identified as sample ‘30.5’, ‘100.7’ and ‘299’. Samples
‘Low level’, ‘Medium level 2’ and ‘High level’ were blind duplicates (BD), and
sample ‘Medium level 1’ was a split level (YP) of Sucralose containing energy-
reduced beverages. The results for duplicate determinations of Sucralose by UV-
absorption and fluorescence are shown in chart form (Mean & Range and Youden
Plots) in Annex II and III.
Three laboratories that participated in the study reported zero or “< XX” values for
some of the test samples.
Laboratory 08 reported several zero and “< XX” values, indicating a consistent and
laboratory inherent bias.
Laboratory 17 reported results from ultraviolet absorption measurement mode at λ =
310 nm only due to lack of equipment to measure fluorescence intensity.
- 12 -
Table 6: Individual results of Sucralose in spiked beverage determined using UV and FL quantification
Sucralose concentration in spiked beverage, [mg/L] 30.5 100.7 299
Lab Code UV UV FL FL UV UV FL FL UV UV FL FL Lab 01 41 33 29 32 69 70 91 67 328 313 330 290 Lab 02 41 53 30 32 116 66 107 86 330 330 328 326 Lab 03 30 32 30 30 111 114 107 102 300 309 308 314 Lab 04 31 45 30 27 110 117 109 113 321 302 284 292 Lab 06 27 55 32 28 119 103 114 93 296 389 285 348 Lab 08 33 58 32 57 105 98 205 0 403 0 Lab 09 36 36 28 32 121 121 112 104 310 328 299 296 Lab 10 28 57 29 61 171 144 299 230 Lab 11 29 52 21 53 60 102 82 116 291 287 293 299 Lab 16 45 43 38 38 106 146 112 119 293 231 277 253 Lab 18 43 26 32 25 118 99 108 65 386 259 352 269 Lab 17 21 43 135 135 279 370
Lab MYC 30 26 31 28 118 127 112 111 289 322 299 300 Data sets in shaded fields were not taken for statistical analysis
Non-compliant data1 Outlier
Table 7: Individual results of Sucralose in energy-reduced beverages determined using UV and FL quantification
Sucralose concentration in energy-reduced beverages, [mg/L] Low level (BD) Medium level (YP) Medium level (BD) High level (BD)
Lab Code UV UV FL FL UV UV FL FL UV UV FL FL UV UV FL FL Lab 01 40 38 72 88 64 64 119 105 104 97 137 127 129 125 Lab 02 43 44 49 42 <58 67 62 66 80 121 85 124 150 102 124 132 Lab 03 46 46 41 46 80 68 77 58 139 117 142 123 117 125 114 115 Lab 04 41 42 40 48 63 58 66 79 125 134 117 116 170 130 127 100 Lab 06 32 39 38 37 97 74 65 80 154 167 147 123 173 130 147 113 Lab 08 <30 67 54 49 <60 83 <60 60 <60 142 <60 131 96 151 114 138 Lab 09 82 57 34 56 74 67 59 137 44 139 139 158 119 131 Lab 10 89 108 58 99 106 102 154 160 154 232 99 131 Lab 11 56 79 60 80 38 68 50 62 117 126 116 136 96 124 98 111 Lab 16 61 49 36 44 97 67 79 58 128 135 86 118 111 137 109 152 Lab 17 51 40 56 114 131 137 139 Lab 18 47 37 44 39 161 45 76 49 121 109 115 101 116 94 108 109
Lab MYC 46 41 48 49 73 73 70 73 120 112 106 112 117 120 113 118 Data sets in shaded fields were not taken for statistical analysis
Non-compliant data1 Outlier
1 Where only 1 set of results was available statistical analysis was not possible, the whole set was rejected as non-compliant data.
- 13 -
Statistical analysis of results
In some cases data was excluded from the statistical analysis, e.g. when statistical
evaluation was impossible (values reported as zero or as “< XX”), or on basis of
problems that were identified by the organiser (e.g. calibration problems).
On basis of the latter case, all results submitted from laboratory 08 were classified as
non compliant prior statistical analysis.
The collaborative trial results were evaluated according to the IUPAC Harmonised
Protocol (7, 8). Outlying results were identified by applying Cochran’s and Grubbs'
tests (p<0.025). Pairs of results that were identified as outliers are indicated with
shaded background in Table 6 and Table 7. The maximum numbers of outliers
identified were two laboratories giving acceptable data ranging from eight to 12
laboratories for ultraviolet absorption and fluorescence measurement mode. Precision
estimates were obtained using the one-way analysis of variance approach according to
the IUPAC Harmonised Protocol (7, 8). Details of the average analyte concentration,
the standard deviations for repeatability (RSDr) and reproducibility (RSDR), the
number of statistical outlier laboratories, the HORRAT ratio and the percentage
recovery for the individual samples are presented in Table 9 at the end of this
document section.
The performance parameters of the method are summarised in Table 8.
- 14 -
Table 8: Method performance parameters obtained in the collaborative trial
Performance parameter Method Measurement
Mode Mean Level
[mg/L] RSDr % RSDR % Recovery %
38 31.4 31.4 124.6 44 11.9 15.0 — 72 20.7 21.5 — 108 16.0 21.6 107.2 124 10.3 14.8 — 130 14.9 16.3 —
UV
312 13.5 13.5 104.3 30 8.9 8.9 98.4 43 15.9 15.9 — 70 14.6 20.7 — 102 15.0 15.6 101.3 119 12.2 17.6 — 119 12.9 12.9 —
HPT
LC
FL
299 9.6 9.6 100 Materials for which no recovery data is given (marked with ‘–‘) were labelled to contain Sucralose.
Comments from collaborative trial participants
Comments were received from three collaborative trial participants. They regarded the
method description as being clear and adequate. Laboratory 02 reported a different
aliquotation scheme used with the intention to save time for calculations. This change
had apparently no effect on the results and was not further regarded for the evaluation
of the end result.
Furthermore several comments were made on the mechanical stability and
homogeneity of the coating of the commercial HPTLC glass plates.
Laboratory 02 reported deviation from the procedure due to instability of the mercury
lamp (too many burning hours) and an urgent need to replace it. For that reason
fluorescence measurement was performed four days after plate development. No
outliers were identified from that lab, indicating that these deviations had no
observable effect on the results while showing the robustness of the method principle.
Laboratory 17 reported deviation from the procedure measuring UV absorption at
wavelength λ = 310 nm. This change had apparently no effect on the results and was
not further regarded for the evaluation of the end result.
- 15 -
Based on the comments from the participants the method description was adjusted at
some points and small editorial additions were made. As the main result, the dilution
procedure was adapted.
Precision characteristics of the method
The precision data for all samples is summarised in Table 9. Based on results for
spiked samples (blind duplicates at three levels), as well as samples taken from the
market and labelled to contain Sucralose (blind duplicates at three levels and one at
split level), the relative standard deviation for repeatability (RSDr) ranged from 10.3 –
32 % for UV detection and from 9.6 – 26 % for FL detection. The relative standard
deviation for reproducibility (RSDR) ranged from 13.5 – 32 % for UV detection and
from 9.6 – 26 % for FL measurement mode.
The recovery values for Sucralose derived from the spiked samples were found to
range from 104.3 – 126.3 % with an average of 112.7 % for ultraviolet detection and
from 99.9 – 100.8 % with an average of 100.3 % for fluorescent detection.
Ultraviolet absorption has previously shown by one of the participants to offer a
detection limit of 5 ng/spot in the plate, resulting in a Sucralose concentration of 1
mg/L in the tested solution, taking into account that 5 µL were applied on the plate (6).
The limit of quantification on the basis of 10x the baseline noise was found as 6 mg/L
and a linear calibration was applicable in the range of 10 – 380 ng of Sucralose for the
ultraviolet absorption measurement mode (6) and 30 – 150 ng/spot for the fluorescent
measurement mode.
Conclusions
The results of this inter-laboratory comparison show satisfying precision characteristics
(RSDr, RSDR and recovery) at the levels of interest that have been stipulated by
European legislation (1) in order to support Commission regulations on legislative
limits. Fluorescence measurement mode turned out to be superior compared with
ultraviolet absorption measurement mode as far as robustness, precision and recovery
were concerned.
- 16 -
HORRAT values are in a few cases slightly higher than 2. This should however not be
considered as unsatisfactory, as the method showed for most levels sufficient
HORRAT values. According to Linsinger and Josephs (9) HORRAT values are also
influenced by the type of method applied and the obtained performance should be seen
with respect to the simplicity of the analytical procedure such as no need for clean-up
and reagent free and simple derivatisation.
Based on the precision values alone, thus recovery, reproducibility and repeatability,
the here described method in both measurement modes can compare with methods for
the determination of food contaminants that qualify for official food control (10).
In conclusion, the method qualified for food beverages at levels equal and above 30
mg/L. The JRC is currently transforming this method into a format suitable for CEN
and will submit it to CEN TC.
Table 9: Results for spiked and test samples, calculated by conventional statistics
UV low (BD) nc2 medium (YP) nc2 medium (BD) nc2 high (BD) nc2 spike 30.5 spike 100.7 Spike 299
# of labs3 8 (3) [2] 8 (4) [1] 10 (3) 11 (2) 12 (1) 11 (2) 11 (2) Mean mg/L 44 72 124 130 38 108 312
Sr mg/L 5.3 14.8 12.7 19.3 11.8 17.3 42.1 RSDr % 11.9 20.7 10.3 14.9 31.4 16.0 13.5 SR mg/L 6.6 15.4 18.3 21.1 11.8 23.4 42.1 RSDR % 15.0 21.5 14.8 16.3 31.4 21.6 13.5
HORRAT4 1.7 2.6 1.9 2.1 3.4 2.7 2.0 Recovery % n.a. n.a. n.a. n.a. 124.6 107.2 104.3
FL low (BD) nc2 medium (YP) nc2 medium (BD) nc2 high (BD) nc2 spike 30.5 spike 100.7 Spike 299
# of labs3 8 (2) [2] 11 (1) 10 (1) [1] 11 (1) 8 (2) [2] 10 (1) [1] 11 (1) Mean mg/L 43 70 119 119 30 102 299
Sr mg/L 6.9 10.2 14.5 15.3 2.6 15.2 28.6 RSDr % 15.9 14.6 12.2 12.9 8.9 15.0 9.6 SR mg/L 6.9 14.4 20.9 15.3 2.6 15.8 28.6 RSDR % 15.9 20.7 17.6 12.9 8.9 15.6 9.6
HORRAT4 1.8 2.4 2.3 1.7 0.9 1.9 1.4 Recovery % n.a. n.a. n.a. n.a. 98.4 101.3 100
n.a. = not applicable; Sr = Standard deviation for repeatability; SR = Standard deviation for reproducibility; RSDr = Relative standard deviation for repeatability (%); RSDR = Relative standard deviation for reproducibility (%); BD = Blind duplicates; YP = Youden matched pair; UV = Ultraviolet detection mode; FL = Fluorescence detection mode
2 nc = (labelled to contain Sucralose) 3 Number in round brackets (-) indicate the number of outliers, numbers in straight brackets [-] indicate non compliant data. 4 HORRAT values are calculated by the modified function, as proposed by Thompson [ref].
References
1. COMMISSION DIRECTIVE 2003/115/EC of 22 December 2003 amending
Directive 94/35/EC on sweeteners for use in foodstuffs. Official Journal of the European Union, 29.01.2004, L 24/65.
2. Kishi, H., K. Kawana (2001) Shokuhin Eiseigaku Zasshi, 42(2): 133-138. 3. Dionex Corporation (2004) Dionex Application Note 159. 4. Stroka, J., N. Dossi, E. Anklam (2003) Food Addit Contam., 20(6): 524-527. 5. Wasik, A., J.McCourt, M.Buchgraber, (2007) J Chromatogr A, 1157(1-2):
187-196. 6. Spangenberg, B., J. Stroka, I. Arranz, E. Anklam (2003) J Liquid Chromatogr.
R.T., 26(16): 2729-2739. 7. Horwitz, W. (1995) Pure & Appl. Chem., 67(2): 331-343. 8. Horwitz, W., R. Albert (1991) JAOAC Int., 74(5): 718-744. 9. Linsinger, T.P.J., R.D. Josephs (2006) Trends in Analytical Chemistry, 25(11): 1125-1130. 10. COMMISSION REGULATION EC/401/2006 of 23 February 2006 laying
down the methods of sampling and analysis for the official control of the levels of mycotoxins in foodstuffs. Official Journal of the European Union, 09.03.2006, L 70/12
Annex Ia
Spiking Protocol
SPIKING PROTOCOL
In order to calculate the recovery of the method each participant will be supplied with 6
different vials marked Spike solution A, Spike solution B, Spike solution C, Spike solution D,
Spike solution E and Spike solution F. These vials contain 3.2 mL spiked sucralose solutions
of unknown concentrations. Store vials at 4 °C. Prior to analysis let them reach room
temperature.
Dilute 1000 µL of spiked solutions B, C, D and F to exactly 10.00 mL (e.g. in a 10 mL
volumetric flask) with methanol-water [50+50 v/v] and shake well to mix thoroughly.
Dilute 2000 µL of spike solutions A and E to exactly 10.00 mL (e.g. in a 10 mL volumetric
flask) with methanol-water [50+50 v/v] and shake well for complete mixing.
These solutions can directly be used for spotting on the NH2-HPTLC plate for determination
of sucralose, no further cleanup is required.
SPIKING PROCEDURE (done in IRMM):
Coke Light (used as ‘blank’) was fortified with sucralose to give three spike solutions (each in
duplicate):
Exactly 1.0 mL, 3.3 mL and 9.8 mL of the sucralose stock standard solution (concentration
6.1 [g/L]) were added to Coke Light to give 200 mL spike solutions with concentrations of 31
[mg/L], 101 [mg/L] and 299 [mg/L].
Annex Ib
Instructions for Participants
VALIDATION OF A HPTLC ANALYTICAL METHOD TO
DETERMINE THE CONTENT OF THE INTENSE
SWEETENER SUCRALOSE IN BEVERAGES
Draft of a method for the determination of sucralose by HPTLC in beverages in a
suitable format for intercomparison purposes.
ABSTRACT
This method for the determination of sucralose in beverages is based on high-performance
thin layer chromatography (HPTLC) in combination with a reagent free derivatisation.
Samples of carbonated and still beverages are diluted with methanol-water and applied
directly on a NH2-HPTLC plate. The plate is developed with an acetonitrile-water mobile
phase and after development directly heated at 190 ºC for 20 minutes. This procedure results
in the formation of UV-light absorbing (and fluorescent) sucralose derivatives. That can be
used for quantification in both modes (UV and FL).
1 SCOPE
This draft specifies a method for the determination of sucralose in beverages using HPTLC.
2 PRINCIPLE
A known quantity of beverage is spotted/sprayed without any clean-up on a NH2-HPTLC
plate. After development plate is directly heated at 190 ºC for 20 minutes to obtain UV-
absorbing and fluorescing sucralose derivatives. The latter are quantitatively determined by
both UV absorption and fluorescence.
3 REAGENTS
Upon receipt store all vials in a refrigerator at 4 °C and allow reach room temperature
before use.
3.1 General
During the analysis, unless otherwise stated, use only reagents of recognized analytical grade
and only distilled water or water of grade 1 as defined in EN ISO 3696. Solvents shall be of
quality for HPLC analysis.
3.2 Methanol
3.3 Acetonitrile
3.4 NH2-HPTLC plates – you will be provided with eight NH2-HPTLC plates 10 x 10 cm
3.5 HPLC syringe filter 13 mm 0.45 µm Nylon – you will be supplied with 5 filters
3.6 Dilution solution – Mix 50 parts per volume of methanol (3.2) with 50 parts per
volume of water (3.1)
3.7 Mobile phase – Mix 80 parts per volume of acetonitrile (3.3) with 20 parts per volume
of water (3.1)
3.8 Sucralose stock solution – Stock solution with mass concentration of 107 [mg/L]. This
solution will be provided for the collaborative trial.
3.9 Working standard solutions for calibration – Pipette amounts of 600 µL, 1300 µL,
2000 µL and 2800 µL of the stock solution (3.8) into different 10 mL volumetric flasks.
Fill the flasks up to the mark with methanol-water (3.6) and shake. This will result in
sucralose solutions with mass concentrations of: 6.4 [mg/L], 13.9 [mg/L], 21.4 [mg/L]
and 30 [mg/L].
3.10 Spike recovery solutions – You will be provided with 6 vials containing spike
solutions of unknown sucralose concentrations.
4 APPARATUS
Usual laboratory equipment and, in particular, the following:
4.1 Syringe of 3 mL – you will be provided with one sterile syringe for single use
4.2 Volumetric flasks of 10 mL
4.3 Spotting or spraying device – (Micro-capillary pipettes of 5 µL can be used instead)
4.4 HPTLC chamber for planar chromatography
4.5 Densitometer (UV) and/or (FL) with Data processing system
4.6 Drying oven capable of maintaining temperature of 190 ± 5 °C for at least 20 minutes
should be used. Alternatively a temperature controlled heating plate can be used
provided the heating surface is equally hot at all locations (thick metal plate).
5 PROCEDURE
5.1 Dilution
Dilute 1000 µL of sucralose test solutions “Sample 01”, “Sample 02”, “Sample 03”, “Sample
05”, “Sample 06”, “Sample 07” and spike solutions B, C, D, F to exactly 10.00 mL (e.g. in a
10 mL volumetric flask) with methanol-water [50+50] and shake well to mix thoroughly.
2
Dilute 2000 µL of test solutions “Sample 04”, “Sample 08” and spike solutions A and E to
exactly 10.00 mL (e.g. in a 10 mL volumetric flask) with methanol-water [50+50] and shake
well for complete mixing.
Shake well turbid solutions (“Sample 02” and “Sample 06”) and after dilution prior to
application filter through 0.45 µm Nylon HPLC syringe filter.
All solutions can directly be used for spotting on the HPTLC plate and determination of
sucralose, no further cleanup is required.
5.2 Application
On a 10 x 10 cm NH2-HPTLC plate apply (by spotting or spraying) each 5 µL of the diluted
solutions in the following manner:
Keep a distance of 20 mm from left and right for the outer spots and of 8 mm from the bottom
edge for the application line.
You will need to spot/spray two test solutions (S1 and S2) and four working standard
solutions for calibration (3.9), resulting in 6 spots per plate. Alternatively band wise
application can be used (e.g. 7 mm-bands).
20 ● ● ● ● ● ● 20
S1 STD1 STD2 STD3 STD4 S2 8 mm
5.3 Preparation of a calibration graph
Prepare calibration graph by applying 5 µL of the four standard solutions proposed in 3.9 on
each HPTLC plate. Plot the peak height values of the sucralose calibration solutions (3.9)
against the concentration of sucralose in [mg/L] in the sample.
5.4 Development
After the solvent has evaporated, develop the plate in either a horizontal or a standard TLC-
chamber for 7 cm with the mobile phase (acetonitrile-water [8+2]).
5.5 Derivatisation
For derivatisation, put the developed plate immediately into a drying oven at 190°C for 20
min., alternatively a temperature controlled heating plate can be used.
5.6 Evaluation
After heating for 20 minutes, allow the plate reach room temperature. If needed, you can
examine the plate first under UV-light (long wavelength filter). Sucralose appears as yellow
fluorescent spot.
3
Scan the plate in the densitometer for UV-absorption at 254 nm, with a slit width of about
120% of the diameter of the spot (for band wise application 80%, e.g. 5 mm slit). Scan also
for fluorescence (FL) at 366 nm with the same slit width.
5.7 Calculation
Quantitative determination of sucralose is carried out by integration both in UV and FL mode
of the peak height. Determine the content of sucralose in the test material, in mg/L, via the
given formula:
Sucralose 21]/[]/[
VVLmgcLmg ×=
Where:
c = concentration in the applied solution calculated from linear regression
V1 = volume of diluted sample solution [10 mL]
V2 = volume of sample taken for dilution [mL]
6 REPORTING
For reporting please use the attached reporting sheet (one per each HPTLC plate) and also
report any comments on the procedure you think are worth mentioning.
4
Annex IIa
Mean & Range Plots for UV-determination
NF 1_low BlindD 04-08.xls mean&range
Page 1
0 : blind replicates
0203
04
06
16
18
17MYC
8, nc
10, GS
11, GD
9, nc
8, nc
1, nc
0
20
40
60
80
100
120
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
NF 2_med YoudP 01-03.xls mean&range
Page 1
0 : split level
01
03
04
06
09
11
16
MYC
8, nc 10, nc
18, C
17, nc
10, nc8, nc
2, nc
0
20
40
60
80
100
120
140
160
180
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
NF 3_med BlindD 02-06.xls mean&range
Page 1
0 : blind replicates
01
02
03 04
06
11
16
18
17
MYC
8, nc
10, nc
9, nc
8, nc0
20
40
60
80
100
120
140
160
180
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
NF 4_high BlindD 05-07.xls mean&range
Page 1
0 : blind replicates
0102
03
04 0609
11
16
18
17
MYC
10, nc
8, nc
0
50
100
150
200
250
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
SP 030.5 AE.xls mean&range
Page 1
0 : blind replicates
01
02
03
04
06
09
10
11
16
18
17
MYC
nc
0
10
20
30
40
50
60
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
SP 100.7 BD.xls mean&range
Page 1
0 : blind replicates
01
02
03 0406
09
11
16
18
17
MYC
10, nc10, nc
8, nc
0
20
40
60
80
100
120
140
160
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
SP 299.0 CF.xls mean&range
Page 1
0 : blind replicates
0102
0304
06
09
11
16
18 17
MYC
8, nc 10, nc10, nc8, nc0
50
100
150
200
250
300
350
400
450
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
Annex IIb
Youden Plots for UV-determination
NF 1_low BlindD 04-08.xls youden
Page 1
0 : blind replicates
MYC 1718
16
0604
0302
11, GD
10, GS
9, nc
8, nc 1, nc0
20
40
60
80
100
120
0 10 20 30 40 50 60 70 80 90 100
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
NF 2_med YoudP 01-03.xls youden
Page 1
0 : split level
MYC
1611
09 06
04
03
01
18, C
17, nc
10, nc8, nc
2, nc
0
20
40
60
80
100
120
140
160
180
0 20 40 60 80 100 120 140 160 180
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
NF 3_med BlindD 02-06.xls youden
Page 1
0 : blind replicates
MYC
17
18
16
11
06
04
0302
01
10, nc
9, nc
8, nc0
20
40
60
80
100
120
140
160
180
0 20 40 60 80 100 120 140 160 180
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
NF 4_high BlindD 05-07.xls youden
Page 1
0 : blind replicates
MYC
17
18
16
11
09
060403
02
01
10, nc
8, nc0
50
100
150
200
250
0 20 40 60 80 100 120 140 160 180 200
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
SP 030.5 AE.xls youden
Page 1
0 : blind replicates
MYC
17
18
16
11
10
09
06
04
03
02
01
nc0
10
20
30
40
50
60
0 5 10 15 20 25 30 35 40 45 50
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
SP 100.7 BD.xls youden
Page 1
0 : blind replicates
MYC
17
18
16
11
09
06
0403
0201
10, nc 8, nc0
20
40
60
80
100
120
140
160
0 20 40 60 80 100 120 140 160
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
SP 299.0 CF.xls youden
Page 1
0 : blind replicates
MYC
17
18
16
11
09
06
0403
02
01
10, nc8, nc0
50
100
150
200
250
300
350
400
450
0 50 100 150 200 250 300 350 400 450
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
Annex IIIa
Mean & Range Plots for FL-determination
NF 1_low BD 04-08.xls mean&range
Page 1
0 : blind replicates
0203 04
06
09
1618
MYC
10, GD
11, GD
nc
nc
0
20
40
60
80
100
120
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
NF 2_med YP 01-03.xls mean&range
Page 1
0 : split level
01 02
03
04 06
09
10
11
16
18
MYC
nc
0
20
40
60
80
100
120
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
NF 3_med BD 02-06.xls mean&range
Page 1
0 : blind replicates
0102
03
04
06
10
11
16
18 MYC
09, C
nc
0
20
40
60
80
100
120
140
160
180
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
NF 4_high BD 05-07.xls mean&range
Page 1
0 : blind replicates
01 02
03 04
06
09
10
11
16
18
MYC
nc
0
20
40
60
80
100
120
140
160
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
SP 030.5 AE.xls mean&range
Page 1
0 : blind replicates
01 0203
0406 09
18MYC
11, C16, GD
ncnc
0
10
20
30
40
50
60
70
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
SP 100.7 BD.xls mean&range
Page 1
0 : blind replicates
01
02
03
04
0609
11
16
18
MYC
10, GSnc
0
50
100
150
200
250
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
SP 299.0 CF.xls mean&range
Page 1
0 : blind replicates
01
02
03
04
06
09
10
11
16
18MYC
nc
0
50
100
150
200
250
300
350
400
450
laboratory number
lab.
mea
n an
d ra
nge,
mg/
kg
accepted dataoutlier datanon compliant dataoverall mean
Annex IIIb
Youden Plots for FL-determination
NF 1_low BD 04-08.xls youden
Page 1
0 : blind replicates
MYC
18
16
0906
0403
02
11, GD
10, GD
nc
nc0
20
40
60
80
100
120
0 10 20 30 40 50 60 70
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
NF 2_med YP 01-03.xls youden
Page 1
0 : split level
MYC
18
16
11
10
09
0604
03
0201
nc
0
20
40
60
80
100
120
0 20 40 60 80 100 120
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
NF 3_med BD 02-06.xls youden
Page 1
0 : blind replicates
MYC
18
16
11
10
06
04
0302
01
09, C
nc
0
20
40
60
80
100
120
140
160
180
0 20 40 60 80 100 120 140 160 180
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
NF 4_high BD 05-07.xls youden
Page 1
0 : blind replicates
MYC
18
16
11
10 09
06
04
03
02
01
nc
0
20
40
60
80
100
120
140
160
0 20 40 60 80 100 120 140 160
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
SP 030.5 AE.xls youden
Page 1
0 : blind replicates
MYC
18
09
0604
03
0201
16, GD
11, C
nc
nc
0
10
20
30
40
50
60
70
0 5 10 15 20 25 30 35 40
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
SP 100.7 BD.xls youden
Page 1
0 : blind replicates
MYC
18
1611
09
06
04
03
02
01
10, GS
nc
0
50
100
150
200
250
0 20 40 60 80 100 120 140 160 180
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
SP 299.0 CF.xls youden
Page 1
0 : blind replicates
MYC
18
16
11
10
09
06
04
03
02
01
nc0
50
100
150
200
250
300
350
400
450
0 50 100 150 200 250 300 350 400 450
Rep 1, mg/kg
Rep
2,
mg/
kg
accepted dataoutlier datanon compliant data45° line
European Commission EUR 23056 EN – Joint Research Centre – Institute for Reference Materials and Measurements Title: Validation of an Analytical Method to Determine the Content of Sucralose in Beverages Author(s): I. Doncheva and J. Stroka Luxembourg: Office for Official Publications of the European Communities 2007 – 62 pp. – 21 x 29,7 cm EUR – Scientific and Technical Research series – ISSN 1018-5593 Abstract An inter-laboratory comparison was carried out to evaluate the performance characteristics of a method for the determination of sucralose in beverages, which was developed at the JRC-IRMM. The method is based on high-performance thin layer chromatography (HPTLC), and reagent-free derivatisation followed by ultraviolet/fluorescence detection. It was tested for the determination of Sucralose (C12H19Cl3O8; (2R,3R,4R,5S,6R)-2-[(2R,3S,4S,5S)-2,5-bis(chloromethyl)-3,4-dihydroxy-oxolan-2-yl]oxy-5-chloro-6-hydroxymethyl)oxane-3,4-diol; CAS No: 56038-13-2) in carbonated and still alcoholic and non-alcoholic beverages at proposed European regulatory limits according to Directive 2003/115/EC (i). Precise determination of Sucralose levels in some of the matrices for which European legislative limits apply, required a robust and reliable analytical method. HPTLC employing reagent-free derivatisation offered such a reliable but simple, fast, cost-effective and environment friendly method (very limited quantities of organic solvents methanol and acetonitrile were used). Separation of Sucralose was performed by direct application of samples (diluted, degassed and/or filtered, if necessary) on amino-bonded silica gel HPTLC plates without prior cleanup and development with acetonitrile:water. The sweetener was determined after heating of the developed plate to 190°C and quantified both in ultraviolet absorption and fluorescence measurement mode. Beverages spiked with sucralose as well as beverages taken from the market and labelled to contain sucralose, were sent to 14 laboratories in five different countries following IUPAC guidelines. A sample that did not contain measureable amounts of sucralose was spiked at levels of 30.5 mg/L, 100.7 mg/L and 299 mg/L. Recoveries ranged from 104 – 125 % with an average of 112 % for ultraviolet detection and from 98 – 101 % with an average of 100 % for fluorescent detection. Based on results for spiked samples (blind duplicates at three levels), as well as samples containing Sucralose (blind duplicates at three levels and one split level), the relative standard deviation for repeatability (RSDr) ranged from 10 – 31 % for ultraviolet detection and from 9 – 16 % for fluorescence detection. The relative standard deviation for reproducibility (RSDR) ranged from 14 – 31 % for ultraviolet detection and from 9 – 21 % for fluorescence detection. The limit of quantification on the basis of 10x the baseline noise was 6 mg/L and response was linear in the range between 30 – 150 ng/spot. The method is therefore considered suitable for the determination of Sucralose in beverages at the proposed European legislative limits.
How to obtain EU publications Our priced publications are available from EU Bookshop (http://bookshop.europa.eu), where you can place an order with the sales agent of your choice. The Publications Office has a worldwide network of sales agents. You can obtain their contact details by sending a fax to (352) 29 29-42758.
The mission of the JRC is to provide customer-driven scientific and technical support for the conception, development, implementation and monitoring of EU policies. As a service of the European Commission, the JRC functions as a reference centre of science and technology for the Union. Close to the policy-making process, it serves the common interest of the Member States, while being independent of special interests, whether private or national.